Publications document the research institute’s explicit, inspectable outputs.
They serve as stable reference points for collaborators, partners, and institutions engaging with our work.
Documents that describe system structure, execution models, and boundary definitions. These texts focus on how components relate and why certain layers must exist.
Formal descriptions of how language, execution, and verification are translated into computable processes.
Analysis of how AI agents operate within institutional structures, including responsibility, escalation, and authority boundaries.
Texts that clarify what the system does not do, where responsibility remains human, and where automation must stop.
Product documentation, announcements, and case studies are published elsewhere. Publications remain stable references, even as implementations evolve.
These texts may be cited in internal design discussions, external collaborations, standards work, and long-term architectural planning.
They do not constitute product commitments or commercial guarantees.